Companionway

Taking Advantage of *args and **kwargs

I consistently forget the option word I have included in a function and so I had to design a way to accept different keywords (option) names to trigger features with in a function.

As an example, assigning a color to a box border is very easy by just adding something like this to a function’s arguments:

def buildabox(msg, border_color="red"):
  """
  docs go here to put a box around msg
  """
  ...

But, if you accidenally call the function like this, it will fail. my_box = buildabox(“my message”, box_color=“white on back”): …

Some thoughts on personal coding workflow

My coding is rudimentary at best but the lessons I have learned along the way save me from constant anguish.

The tool of choice is mogrify which come from the imagemagic package on your linux platform.

I had one image that looked fine when I displayed it using display tools in linux but when placed into a webpage the image would be reversed.

Python all() function

Much of my coding efforts are directed towards analyzing financial data specifically targeting potential cash secured puts or covered calls. Recently, I had to refine some market overview displays. I needed a listing of the current market “gainers” and market “losers”. There is one site which I scrape using panda read_html function to grab a hadful of tables. My problem arises from two factors. One is there are no labels or titles to the tables and they can vary. The a table can arise from special (temporart) events and the weekends or holidays might call for the tables to re-arrange their order. This requires that I look directly at the panda table and derive which table I might be examining. One table I look for lists all the “gainers” currenly in the market. What I do know is that the third column of that table will always contain a plus sign (the % change of all the gainers). I had nested loops initially to do this but I refactored the code to use the “all()” function to discover which table has all “+” signs in the column 3. My standard caveat applies here - there is likely a better way to do this - so let me know.

Our last post discussed using fisa-vimrc and appreciating the simple installation and ease of use. This post will explore, briefly, some tweaks and a short “how to” for each of the loaded plugins. This exercise will be worth your time if you have not explored these tools before.

I have posted several articles about python fabric and yet I failed to warn you about the dangers of version and module chaos.

Initially, fabric came as a python2 only application. They added support for python3 later. Here is the version I am using in a virtual environment (more about that in a minute). From within my pipenv project directory with pipenv shell loaded this is what I get running fab –version

fab --version
Fabric3 1.14.post1
Paramiko 2.4.2

Here is how I created my virtual environment. I have used almost all of python’s virtual environment tools (and there a quite a few). And there has been a steady progression of improvements in those tools. My current goto for a python virtual environment is pipenv. It is simple, clean, and almost fool proof which is essential for me.

Python fabric is a library tool for executing ssh commands remotely and responding as you desire. I use python fabric for promoting code and web content. Essentially it is my devops tool to satisfy the single most important devops rule: never log into a server! I use this over ansible or some of the other devop tools because fabric only requires knowning one langauge and that is python.

My network changes a lot as I take servers down or stand them up frequently. Normally that would require changing the environment roles (host lists) in my fabfile.py with all my changes. In the past I set aside certain IPs in my network for each role. For example all my raspberry pies get named before I ever introduce them to the network. Using the option –skip_bad_hosts with fabric is one way to do this but I really want more control.